Abstract
The application of high-bandwidth networks and cloud computing in manufacturing systems will be followed by mass data. Industrial data analysis plays important roles in condition monitoring, performance optimization, flexibility, and transparency of the manufacturing system. However, the currently existing architectures are mainly for offline data analysis, not suitable for real-time data processing. In this paper, we first define the smart factory as a cloud-assisted and self-organized manufacturing system in which physical entities such as machines, conveyors, and products organize production through intelligent negotiation and the cloud supervises this self-organized process for fault detection and troubleshooting based on data analysis. Then, we propose a scheme to integrate knowledge reasoning and semantic data where the reasoning engine processes the ontology model with real time semantic data coming from the production process. Based on these ideas, we build a benchmarking system for smart candy packing application that supports direct consumer customization and flexible hybrid production, and the data are collected and processed in real time for fault diagnosis and statistical analysis.
Highlights
Smart factories and smart production are drawing increasing attention from both the academic community and industrial pioneers [1]
The high dynamics comes the intelligent negotiation that occurs from one operation to another; during negotiation, the from the intelligent negotiation that occurs from one operation to another; during negotiation, resources are determined in real time instead of based on pre-allocation
In the implementation of smart factories, we apply several emerging technologies such as Ethernet network, cloud computing, big data, and artificial intelligence to achieve the vertical integration industrial Ethernet network, cloud computing, big data, and artificial intelligence to achieve the of automation systems with information systems
Summary
Smart factories and smart production are drawing increasing attention from both the academic community and industrial pioneers [1]. Big data analytics plays an essential role in a smart factory It acts like a brain, extracting valuable knowledge to serve decision-making in different layers to support applications such as global performance optimization, supervisory control, and fault diagnosis, as well as proactive operation and maintenance [8,9]. We present a framework for smart factories to arrange involved components into a layered and structural architecture, based on which we could identify different data processing hosts and separating online real-time processing from offline batch processing. We use a smart factory prototype, simulating a personalized candy packing application, to verify the proposed architecture and method.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
